Re: [ccp4bb] Refmac and hydrogens -developers

[James Holton]

The first thing you want to do when debugging hydrogens in refmac5 is to 
make sure it outputs whatever it was doing:

MAKE HOUT Y

This is not the default.  Normally, refmac will take a PDB file without 
any hydrogens in it, quietly add them before doing refinement, and then 
delete them before outputting the final coordinates.  Why so secretive?  
Probably because someone complained about having to look at all that 
fuzzy hydrogen nonsense in the graphics.  In fact, I know some 
crystallographers who refuse to believe that the element hydrogen even 
exists. Never seen it. Including them in refinement has even been called 
"too many parameters". Nowadays, of course, we know having hydrogens 
involved in the geometry is a good thing.  This is especially true at 
poor resolution.


The other important hydrogen keyword to know is:
MAKE HYDR A

This is the default.  "A" means add all hydrogens, regardless of what is 
in the input pdb file.  Normally, these get deleted before output, so 
you don't notice it.  The other choices are:
MAKE HYDR N: which is basically "do not do the hydrogen thing".  Any 
hydrogens in the input file will be ignored, and they will also not be 
in the output file (even if you use MAKE HOUT Y), because they were 
never there.
MAKE HYDR Y : this is the one that will use the hydrogens listed in the 
pdb file.  Here there lies a danger!  Not every program names hydrogens 
the same way, so some might be quietly ignored.  The only way to know if 
they were is to use MAKE HOUT Y.


So, if you want to define and use your own hydrogens through many rounds 
of refmac, you want to use:

MAKE HYDR Y
MAKE HOUT Y

But, if your starting PDB has no hydrogens in it you want your first 
round to have:

MAKE HYDR A
MAKE HOUT Y

If you inspect the occupancies of the output hydrogens you may find that 
some are zero, and every once ina while one is somewhere between 1 and 
0.  I have not seen documentation for this, but it appears that refmac 
might be quietly doing some sort of occupancy refinement on hydrogens 
that may or may not be there.  Titratable ones for example.  The 
philisophy explained to me once by Garib is that since it is twice as 
bad to put in an electron that isn't there vs leaving out one that is 
you want to err on the side of dropping hydrogens that don't seem to fit.


HTH,

-James Holton
MAD Scientist


On 5/20/2020 7:07 PM, Bernhard Rupp wrote:


Hi Fellows,

for an experiment, I am running 0.9 A data with unrestrained Refmac 
(yes I know should/could use SHELXL, but let’s drop that for now).


When I select ‘ignore even if present in file’ in a PDB that has 
hydrogens, I get the identical results than with ’use if present’ or 
‘generate all’.


The log informs me that the instructions were properly issued, the 
output PDB does not have Hs, but Rs and map are exactly the same


as with selection of if present or generate all H. Does not seem to 
make sense.


If I cull the hydrogens from the input PDB and ‘use if present’ or 
‘ignore’, the stats and map are different and no H in output as 
requested -  all as expected.


Maybe that can be reproduced and, if it is not a feature, fixed.

Best, BR

--

Bernhard Rupp

http://www.hofkristallamt.org/

b...@hofkristallamt.org 

+1 925 209 7429

+43 676 571 0536

--

Real knowledge is to know

the extent of one's ignorance

--




To unsubscribe from the CCP4BB list, click the following link:
https://www.jiscmail.ac.uk/cgi-bin/webadmin?SUBED1=CCP4BB=1






To unsubscribe from the CCP4BB list, click the following link:
https://www.jiscmail.ac.uk/cgi-bin/webadmin?SUBED1=CCP4BB=1

This message was issued to members of www.jiscmail.ac.uk/CCP4BB, a mailing list 
hosted by www.jiscmail.ac.uk, terms & conditions are available at 
https://www.jiscmail.ac.uk/policyandsecurity/

Re: [ccp4bb] Refmac and hydrogens -developers

[Eleanor Dodson]

Yes - I am having problems with hydrogens in REFMAC recently - has
something changed?
My problem is this: Build a rough model for some residues into a lousy map.
(R ~ 32% etc..)
I know it is rough with clashes and bad geometry but trust refinement to
improve things a bit..
However if the run includes instruction BUILD hydrogens the structure sort
of explodes - R up to 39%.
If I dont do that REFMAC does fix things a bit - R drops a bit - ready for
more corrections..

I assume that now the hydrogens are generating restraints and these
restraints are blowing things up.

Surely hydrogens should not be allowed to do this?
Eleanor


On Thu, 21 May 2020 at 03:07, Bernhard Rupp 
wrote:

> Hi Fellows,
>
>
>
> for an experiment, I am running 0.9 A data with unrestrained Refmac (yes I
> know should/could use SHELXL, but let’s drop that for now).
>
>
>
> When I select ‘ignore even if present in file’ in a PDB that has
> hydrogens, I get the identical results than with ’use if present’ or
> ‘generate all’.
>
>
>
> The log informs me that the instructions were properly issued, the output
> PDB does not have Hs, but Rs and map are exactly the same
>
> as with selection of if present or generate all H. Does not seem to make
> sense.
>
>
>
> If I cull the hydrogens from the input PDB and ‘use if present’ or
> ‘ignore’, the stats and map are different and no H in output as requested
> -  all as expected.
>
>
>
> Maybe that can be reproduced and, if it is not a feature, fixed.
>
>
>
> Best, BR
>
> --
>
> Bernhard Rupp
>
> http://www.hofkristallamt.org/
>
> b...@hofkristallamt.org
>
> +1 925 209 7429
>
> +43 676 571 0536
>
> --
>
> Real knowledge is to know
>
> the extent of one's ignorance
>
> --
>
>
>
> --
>
> To unsubscribe from the CCP4BB list, click the following link:
> https://www.jiscmail.ac.uk/cgi-bin/webadmin?SUBED1=CCP4BB=1
>



To unsubscribe from the CCP4BB list, click the following link:
https://www.jiscmail.ac.uk/cgi-bin/webadmin?SUBED1=CCP4BB=1

This message was issued to members of www.jiscmail.ac.uk/CCP4BB, a mailing list 
hosted by www.jiscmail.ac.uk, terms & conditions are available at 
https://www.jiscmail.ac.uk/policyandsecurity/

[ccp4bb] Refmac and hydrogens -developers

[Bernhard Rupp]

Hi Fellows,

 

for an experiment, I am running 0.9 A data with unrestrained Refmac (yes I
know should/could use SHELXL, but let's drop that for now).

 

When I select 'ignore even if present in file' in a PDB that has hydrogens,
I get the identical results than with 'use if present' or 'generate all'.

 

The log informs me that the instructions were properly issued, the output
PDB does not have Hs, but Rs and map are exactly the same 

as with selection of if present or generate all H. Does not seem to make
sense.

 

If I cull the hydrogens from the input PDB and 'use if present' or 'ignore',
the stats and map are different and no H in output as requested -  all as
expected.

 

Maybe that can be reproduced and, if it is not a feature, fixed.

 

Best, BR

--

Bernhard Rupp

  http://www.hofkristallamt.org/

  b...@hofkristallamt.org

+1 925 209 7429

+43 676 571 0536

--

Real knowledge is to know 

the extent of one's ignorance 

--

 




To unsubscribe from the CCP4BB list, click the following link:
https://www.jiscmail.ac.uk/cgi-bin/webadmin?SUBED1=CCP4BB=1

This message was issued to members of www.jiscmail.ac.uk/CCP4BB, a mailing list 
hosted by www.jiscmail.ac.uk, terms & conditions are available at 
https://www.jiscmail.ac.uk/policyandsecurity/